Traditionally, admission control is only considered frequently if there is sufficient capacity on the destination server. Provisioning or other tasks in a cloud environment are complex, and there are many complex tools integrated together to automate the operations. This requires integration of several tools and components within the tools, interacting with each other, and passing requests from the user interface down to the control manager tools. The components of a tool may run on different server or on the same server. In highly automated many-step processes also, management system is the cause for majority of problem tickets. For example, automated information technology (IT) management systems may be utilized in running cloud computing and data centers. Automation of many complex integrations create potential for many failure points.
For instance, for a virtual machine (VM) to be provisioned in a virtual environment, or cloud environment, there are many requirements which need to be fulfilled for a request to be satisfied, such as updating different tools, issuing different requests between various management tools which can be located on different physical servers, working network and authentication/valid credentials between servers and running tools, and available resources within each tool. Hence also, there may be a potential for many points of failures. The interaction between tools can be broken due to multiple reasons; For example, if network connection between the tools is not operational, if credentials for authentication between different tools and/or servers expired, critical process is down, and/or there are corrupted configurations, the request will fail, even if there are sufficient resources on the target server to provision a VM.